Epstein-Barr virus enhances genome maintenance of Kaposi sarcoma-associated herpesvirus

Abstract

Primary effusion lymphoma (PEL) is a B cell lymphoma that is always associated with Kaposi's sarcoma-associated herpesvirus (KSHV) and in many cases also with Epstein-Barr virus (EBV); however, the requirement for EBV coinfection is not clear. Here, we demonstrate that adding exogenous EBV to KSHV⁺ single-positive PEL leads to increased KSHV genome maintenance and KSHV latency-associated nuclear antigen (LANA) expression. To show that EBV was necessary for naturally coinfected PEL, we nucleofected KSHV⁺/EBV⁺ PEL cell lines with an EBV-specific CRISPR/Cas9 plasmid to delete EBV and observed a dramatic decrease in cell viability, KSHV genome copy number, and LANA expression. This phenotype was reversed by expressing Epstein-Barr nuclear antigen 1 (EBNA-1) in trans, even though EBNA-1 and LANA do not colocalize in infected cells. This work reveals that EBV EBNA-1 plays an essential role in the pathogenesis of PEL by increasing KSHV viral load and LANA expression.

Keywords: EBV; KSHV; Kaposi sarcoma; LANA; primary effusion lymphoma.

Fig. 1.

EBV and KSHV copy numbers in PELs and their EBV-GFP sublines by digital qPCR. (A–D) Scatter plots from digital PCR analysis show signal from VIC reporter dye on the x axis against signal from fluorescein amidite (FAM) reporter dye on the y axis. Data points represent individual PCR reactions and are color-coded for reporter dye signals: blue for FAM, red for VIC (ERV-3), green for FAM and VIC, yellow for no amplification. (A) Namalwa cells were used as positive control for EBV (FAM: green and blue). (B and C) Signal from EBV (FAM: green and blue) is completely absent in BC3 cells serving as negative control; only the single-copy human gene (ERV-3: red) is visible (B). This is in contrast to the BC3 EBV-GFP sorted and selected subline (C). (D) Signal from the KSHV FAM-labeled probe. (E and F) EBV (E) and KSHV (F) copy numbers in BC3 (red dots) and CRO-AP6 (green dots) cells and their EBV-GFP sublines (blue dots for CRO-AP6 EBV clone 3). Total DNA was analyzed by digital qPCR; values are from two biological replicates per cell line. Gray boxes represent the 25th to 75th percentiles of the data. Data were analyzed by ANOVA adjusted for multiple comparisons by Tukey method. ***P < 0.01.

Fig. 2.

Representative example of a 3D-IFA image created by Imaris software. (A–D) Multipage TIFF format, which encloses 50 individual optical sections from the same sample in a single file, of a BC3 sorted G418⁺ clone. A–C show monochrome, conventionally captured images of single-channel signals from LANA, β-actin, and EBNA-1, respectively, in which all the signals overlap and bleed into each other. (D) In the merged image LANA (labeled in blue) localizes within the nucleus of the cells in foci in characteristic dots; EBNA-1 (in magenta) is labeled to confirm the presence of EBV-GFP after infection. β-Actin is labeled in green. (E and F) 3D images created from the multipage file after signal processing. The signal from LANA was maintained in dots; the signal from EBNA-1 has also been represented as dots to facilitate counting of LANA dots and to provide a clearer view of possible colocalization between the two proteins. (Scale bars: 5 µm.)

Fig. 3.

LANA expression is significantly increased in naturally KSHV and EBV coinfected PEL lines and in the EBV-GFP–sorted G418⁺ cells. (A) The number of LANA dots per cell is shown on the vertical axis. They are significantly increased in BC1 and JSC1 cells, which are naturally EBV and KSHV coinfected, in comparison with BC3 and CRO-AP6 cells, which do not carry EBV. Data were analyzed by ANOVA. ***P < 0.001. (C and E) The number of LANA dots per cell is shown on the vertical axis for BC3 (C) and CRO-AP6 (E) parental cells and after sorting and maintenance in either G418⁺ or G418⁻ medium. Data were analyzed by ANOVA. ***P < 0.001. (B and D) LANA protein levels were ascertained by Western blotting using anti-LANA and anti-actin antibody. +, sorted G418⁺ cells; −, sorted G418⁻ cells; u, unsorted cells. An EBV⁺, KSHV⁻ LCL cell line was used as negative control. LANA dot analysis was performed by counting the dots of more than 100 cells for each cell line using 3D IFA images; for EBV-GFP sublines, only one clone was used for each cell line (clone 6 for BC3 and clone 2 for CRO-AP6).

Fig. 4.

EBV-specific CRISPR/Cas9 nucleofection leads to a decrease in KSHV copy number, LANA expression, and cell viability in naturally coinfected PEL cells. (A) KSHV and EBV viral load as determined by digital PCR is shown on the vertical axis for BC1, JSC1, BC3 (KSHV only), and Namalwa (EBV only) cells at 24 h after nucleofection with the indicated plasmids. Individual values are shown for three independent experiments (a, b, c). Gray boxes represent the 25th to 75th percentiles of the data. (B and C) Western blotting for LANA protein in JSC1 (B) and BC1 (C) cells harvested 48 h after nucleofection with indicated plasmids. β-Actin was loaded for normalization. The BJAB cell line was used as a negative control. Also indicated are molecular weight markers. (D–G) Relative cell viability at 24, 48, and 72 h after nucleofection for EBV⁻ BC3 (D), KSHV⁻ Namalwa (E), EBV⁺KSHV⁺ JSC1 (F) and BC1 (G) cells. Values are measured in log10 RLU of CTG reagent. Individual data for three independent biological experiments (indicated by symbol shape), each with multiple technical replicates, are shown with boxplot of the 25th to 75th percentiles of the data.

Fig. 5.

(A) KSHV and EBV genome copy number is shown on the vertical axis for sgEBV1, -2, and -6 or sgEBV4 and -5 or GFP- or mock-nucleofected cells either alone or with pCEP4. Individual values for two experiments are shown. (B) Western blots for LANA protein expression. β-Actin was loaded for normalization. The BJAB cell line was used as a negative control. (C) Cell viability assay performed by CTG at 24 (yellow), 48 (orange), or 72 (red) h after pCEP4 or mock nucleofection. Values for three independent experiments are shown. The y axis represents log10 RLU.